Skip to main content
Scour
Browse
Getting Started
Login
Sign Up
You are offline. Trying to reconnect...
Copied to clipboard
Unable to share or copy to clipboard
Context Windows
🪟 Context Windows
Specific
Long Context Models, Memory Management, Attention Patterns
Filter Results
Timeframe
Fresh
Past Hour
Today
This Week
This Month
Feeds to Scour
Subscribed
All
Scoured
122
posts in
5.6
ms
Efficient and Training-Free Single-Image Diffusion
Models
🎯
Fine-tuning
haojunqiu.github.io
·
5d
5 days ago
·
Hacker News
Actions for Efficient and Training-Free Single-Image Diffusion Models
End-to-End
Context
Compression at
Scale
🤖
Transformers
Content type:
Academic
arxiv.org
·
2d
2 days ago
Actions for End-to-End Context Compression at Scale
High Bandwidth
Flash
| A New
Memory
for AI Data Centers and Edge Computing | Sandisk
🤖
LLM
ncnonline.net
·
2d
2 days ago
Actions for High Bandwidth Flash | A New Memory for AI Data Centers and Edge Computing | Sandisk
Issue #390 - The ML Engineer 🤖
🤖
AI
Content type:
News
Content type:
Blog
machinelearning.substack.com
·
3d
3 days ago
·
Substack
Actions for Issue #390 - The ML Engineer 🤖
OpenCV 5 release - New DNN engine with enhanced ONNX and
LLM/VLM
support, Intel, Arm, and RISC-V hardware optimizations - CNX Software
👁️
Computer Vision
Content type:
News
cnx-software.com
·
1d
1 day ago
Actions for OpenCV 5 release - New DNN engine with enhanced ONNX and LLM/VLM support, Intel, Arm, and RISC-V hardware optimizations - CNX Software
BeeLlama.cpp DFlash on Strix Halo: 2.7x Gemma 31B, But MTP Is Still Faster
⚡
Inference Optimization
sleepingrobots.com
·
4d
4 days ago
Actions for BeeLlama.cpp DFlash on Strix Halo: 2.7x Gemma 31B, But MTP Is Still Faster
RKSC: Reasoning-Aware
KV
Cache
Sharing and Confident Early Exit for Multi-Step
LLM
Inference
🤖
LLM
Content type:
Academic
arxiv.org
·
1d
1 day ago
Actions for RKSC: Reasoning-Aware KV Cache Sharing and Confident Early Exit for Multi-Step LLM Inference
How LLMs Actually Work: A Friendly Map for Humans • oreoro
🤖
LLM
oreoro.github.io
·
5d
5 days ago
·
Hacker News
Actions for How LLMs Actually Work: A Friendly Map for Humans • oreoro
Benchmarking dots.tts on Strix Halo
🤖
AI
sleepingrobots.com
·
3d
3 days ago
Actions for Benchmarking dots.tts on Strix Halo
Gated DeltaNet, From First Principles
✍️
Prompt Engineering
Content type:
Blog
sankalp.bearblog.dev
·
1d
1 day ago
Actions for Gated DeltaNet, From First Principles
How to cut the cost of
long
AI agent threads (without making the agent dumber)
🤖
Agent
Content type:
Blog
viktor.com
·
2d
2 days ago
·
Hacker News
Actions for How to cut the cost of long AI agent threads (without making the agent dumber)
#065 - Claude writes 80% of Anthropic's own code, Cloudflare buys Vite, ChatGPT ships Dreaming
memory
🔓
Open Source
indiehacker.news
·
6d
6 days ago
Actions for #065 - Claude writes 80% of Anthropic's own code, Cloudflare buys Vite, ChatGPT ships Dreaming memory
Reality: The Final Eval — Lukas Petersson and Axel Backlund of Andon Labs
🤖
Agent
latent.space
·
6d
6 days ago
·
Hacker News
Actions for Reality: The Final Eval — Lukas Petersson and Axel Backlund of Andon Labs
Anatomy of a high-performance EP kernel
🤖
LLM
Content type:
Blog
fergusfinn.com
·
1d
1 day ago
·
Hacker News
Actions for Anatomy of a high-performance EP kernel
JeevanJoshi2061/titan_engine_core:
Constant-memory
sequence
modeling
engine combining selective holographic-compression (ASH-C) with a coordinate pointer network (HEP-DNA). Bypasses the linear
KV
Cache bottleneck on consumer GPUs.
🤖
LLM
Content type:
Code
github.com
·
5d
5 days ago
·
Hacker News
Actions for JeevanJoshi2061/titan_engine_core: Constant-memory sequence modeling engine combining selective holographic-compression (ASH-C) with a coordinate pointer network (HEP-DNA). Bypasses the linear KV Cache bottleneck on consumer GPUs.
FlashMemory-DeepSeek-V4: Lightning Index
Ultra-Long
Context
via Lookahead Sparse
Attention
🤖
LLM
Content type:
Academic
arxiv.org
·
2d
2 days ago
·
Hacker News
Actions for FlashMemory-DeepSeek-V4: Lightning Index Ultra-Long Context via Lookahead Sparse Attention
How the UK Is Turning Sovereign AI Ambition Into Action With NVIDIA Technologies
🎮
Reinforcement Learning
Content type:
Blog
blogs.nvidia.com
·
3d
3 days ago
Actions for How the UK Is Turning Sovereign AI Ambition Into Action With NVIDIA Technologies
DeepSeekV4 1.6T Day 0 to Day 43 Performance Over Time - Huawei, GB300 NVL72, MI355X, B200
🤖
AI
Content type:
News
newsletter.semianalysis.com
·
1d
1 day ago
·
Hacker News
Actions for DeepSeekV4 1.6T Day 0 to Day 43 Performance Over Time - Huawei, GB300 NVL72, MI355X, B200
The iPhone’s Last Stand
🤖
Agent
stratechery.com
·
1d
1 day ago
·
Hacker News
Actions for The iPhone’s Last Stand
Still: Amortized
KV
Cache
Compaction in a Single Forward Pass
⚡
Inference Optimization
Content type:
Academic
arxiv.org
·
2d
2 days ago
Actions for Still: Amortized KV Cache Compaction in a Single Forward Pass
Sign up or log in to see more results
Sign Up
Login
« Page 2
Log in to enable infinite scrolling
Keyboard Shortcuts
Navigation
Next / previous item
j
/
k
Open post
o
or
Enter
Preview post
v
Post Actions
Love post
a
Like post
l
Dislike post
d
Undo reaction
u
Save / unsave
s
Recommendations
Add interest / feed
Enter
Not interested
x
Go to
Home
g
h
Interests
g
i
Feeds
g
f
Likes
g
l
History
g
y
Changelog
g
c
Settings
g
s
Browse
g
b
Search
/
Pagination
Next page
n
Previous page
p
General
Show this help
?
Submit feedback
!
Close modal / unfocus
Esc
Press
?
anytime to show this help